ManhTin / ba-image-classifier-test

Script for running the trained modell for parking space monitoring of my bachelor thesis project.

BA Image Classifier Test

Contains script to run the model created in Training.

Dependencies

• Python 3.7.9
• Pip
• Pipenv pip install pipenv

Installation

Install packages for local development pipenv install --dev

Install packages for deployment

• Install packages pipenv install
• Launch virtual environment where the packages are installed pipenv shell
• To install TensorFlow Lite Interpreter follow TFLite Python quickstart guide for your device

Setup

• Unzip images.zip into project root folder

Run script

• Launch the virtual environment created in the Installation step pipenv shell
• Run script via python main.py or python3 main.py depending on your Python installation
• To exit virtual environment press CTRL + d or type exit

Test Data

The patches inside images.zip were created using patches_generator.py and sorted manually into the folders empty and occupied. You can do it yourself by unzipping images.zip and running python patches_generator.py inside the virtual env. Delete the contents of the folders empty and occupied. Sorting has to be done manually.